US 12268664

I'll search for authoritative information on this patent number across USPTO and litigation sources.

Let me look for the patent's claims text and any Federal Circuit activity.

Let me check the certificate of correction and any Federal Circuit activity specific to this patent.

US Patent 12,268,664 — Summary

Note on scope of my search: I ran targeted searches against Google Patents (full text), USPTO/PTAB records, FDA Orange Book aggregations, and district-court filings. I did not find any CAFC (Federal Circuit) 2026 docket naming this patent. I found no results at all for the certificate-of-correction query. Details below, with uncertainty flagged explicitly.

1. Bibliographic data

Sources: Google Patents, PTAB EX2077 (patent front page)

Continuation chain (per the patent's own Cross-Reference section and the PTAB exhibit): 18/782,666 ← 18/062,062 (Dec 6, 2022) ← 17/228,514 (Apr 12, 2021, now US 11,793,786) ← 15/955,397 (Apr 17, 2018) ← 12/413,439 (Mar 27, 2009) ← provisional 61/040,558 (Mar 28, 2008). Note the PTAB exhibit marks 15/955,397 and 12/413,439 as abandoned; the Google Patents text I was given does not carry that annotation.

2. Abstract (verbatim)

"The invention relates to pharmaceutical compositions comprising one or more benzodiazepine drugs for nasal administration, methods for producing and for using such compositions."

3. Plain-language overview

The patent describes a water-free (or near-water-free) nasal spray solution of a benzodiazepine — principally diazepam — formulated in a carrier of:

• Vitamin E (tocopherol/tocotrienol) as the main solvent (the "oil" phase), and
• a short-chain alcohol or glycol (e.g., ethanol, benzyl alcohol) as a co-solvent to cut viscosity and improve sprayability.

The stated problem (Background): oral benzodiazepines are slow and suffer first-pass metabolism; IV requires trained professionals and sterile technique; suppositories are impractical for a bystander. The stated advantage is a quickly administered, needle-free rescue medicine that can be given by a caregiver (or the patient) during a seizure or during a pre-seizure "aura," with claimed onset in some instances under 30, 15, 10, or even 5 minutes.

The specification is broad and mostly prophetic: it lists many benzodiazepines (alprazolam, diazepam, flurazepam, lorazepam, medazepam, mexazolam, midazolam, temazepam, etc.), discusses both solutions (drug fully dissolved) and particulate suspensions (microparticles/nanoparticles, <~5000 nm), and includes alkyl glycoside penetration enhancers (including the Intravail® class and dodecyl β-D-maltoside). Notably, for the solution embodiments the specification excludes Vitamin E TPGS, while for particulate suspensions it teaches that Vitamin E TPGS is a desirable stabilizer.

Important gap: although the abstract and description recite "methods for producing and for using" and the specification contains method-of-treatment embodiments, the claims as reported by secondary sources appear to be composition claims only (drugpatentwatch lists the claim type as "Composition").

4. Independent claim (plain language)

Per drugpatentwatch and the D. Del. complaint (the two sources agree verbatim), there is one independent claim, claim 1:

"1. A pharmaceutical composition for intranasal administration comprising:

• about 1 to about 20 mg of diazepam dissolved in
• about 45% to about 85% (w/w) of one or more natural or synthetic tocopherols or tocotrienols selected from α-, β-, γ-, δ-tocopherol, α-, β-, γ-, δ-tocotrienol, tocophersolan, any esters thereof, and combinations thereof; and
• about 25% to about 40% (w/w) of one or more alcohols selected from ethanol, propyl alcohol, butyl alcohol, pentanol, benzyl alcohol, and combinations thereof; and
• about 0.01% to about 1% (w/v) of dodecyl β-D-maltoside;

wherein the composition is a solution and contains less than 1% water, and wherein the composition is in a pharmaceutically-acceptable spray formulation."

In plain terms, claim 1 covers: a nasal-spray liquid in which 1–20 mg of diazepam is fully dissolved (not suspended) in a Vitamin E/alcohol solvent system that is at least 99% free of water, with a specific maltoside (dodecyl β-D-maltoside) absorption enhancer present at 0.01–1% w/v.

Claims 2–18 appear to be dependent on claim 1 and narrow it, e.g.:

• claim 2: the tocopherol is specifically α-tocopherol;
• claims 3–4: diazepam concentration about 10–250 mg/mL, or about 20–50 mg/mL;
• claim 5: diazepam about 2–10 mg;
• claim 6: tocopherol at about 60–75% (w/w);
• claim 7: additional excipients (stabilizers, colorants, pH adjusters, buffers, preservatives, wetting agents, flavoring agents);
• claims 8–10: fill volume about 25–250 µL, 50–150 µL, or about 100 µL;
• claims 11–14: progressively lower water limits — <0.5%, <0.25%, <0.1%, and no water;
• claims 15–16: no more than 0.01% of a diazepam impurity, including across stability conditions 25 °C/60% RH, 30 °C/65% RH, and 40 °C/75% RH;
• claim 17 (and beyond): spray content uniformity stability characteristics.

Uncertainty: the patent has 18 claims, and my secondary-source listing was truncated at claim 17 (mid-sentence). I could not verify from an authoritative full-text source whether there is a second independent claim (e.g., claim 18) or whether claims 17–18 are further dependents. All claims I could see were composition claims depending from claim 1. Treat the independent-claim analysis as high-confidence for claim 1, incomplete for any claim beyond 17.

5. Post-issue activity and litigation

• Certificate of Correction: A post-issue communication / certificate of correction for US12268664 is recorded as dated January 22, 2026 (GreyB/Pharsight). I could not retrieve the substance of that correction — my searches returned no document text. Because a certificate of correction can change claim wording, the claim text above (sourced from secondary databases and a 2025 complaint) should be verified against the corrected printed claim before relying on it. Flagging this as material uncertainty.
• PTAB: IPR2025-00464, caption Padagis US LLC v. Neurelis, Inc. The '664 patent appears in that proceeding as Patent Owner Exhibit EX2077 (filed Feb. 20, 2026). One document header renders the case as involving "U.S. PATENT 8,895,546" while the exhibit itself is the '664 patent — I am reporting that literally rather than reconciling it.
• District court (D. Del.): The Google Patents record lists 1:25-cv-00821 and 1:26-cv-00261 for the '664 patent family (family ID 41114811). Related complaints (Neurelis v. Padagis, 1:26-cv-00401, filed Apr. 8, 2026) assert the '664 patent against Padagis's ANDA No. 219320 and Lupin's ANDA No. 220394 for generic Valtoco. The complaints contrast the "comprising" language of the '664 patent with the "consisting of" language of the '414 and '786 patents, and identify construction of "about" in the weight-percentage ranges as a central issue. Sources: paragraphfour.com complaint PDF, ExParte complaint analysis, CourtListener 1:26-cv-00401
• Trade name: the '664 patent is one of ~7–12 Orange Book patents listed against VALTOCO (diazepam nasal spray, 5/7.5/10 mg, NDA 211635, approved Jan 10, 2020).

6. Explicit uncertainties

1. No CAFC 2026 docket found for US 12,268,664. The Federal Circuit–adjacent activity I surfaced concerns Aquestive/Libervant (a D.C. Circuit matter, Neurelis v. Califf), not the '664 patent and not the CAFC. Absence of a CAFC hit is not proof none exists.
2. Certificate of correction dated 2026-01-22 exists but its content is unknown to me; claims could have changed.
3. Claim 18 is unverified (truncated listing).
4. The Google Patents full text provided to me contains the description and abstract but does not include the claims section, so my claim analysis rests on the Delaware complaint and drugpatentwatch, which agree verbatim on claim 1.
5. I am treating the expiry as March 27, 2029 (from the 2009 parent filing), consistent with both Google Patents and Orange Book aggregators; this reflects the terminal-disclaimer indication on the face of the patent.

I'll search for litigation involving this specific patent number.

Let me dig into the specific cases and the Unified Patents portal.

Let me get the patent-specific litigation page and the docket details.

Bottom line

Yes — US 12,268,664 ("the '664 patent") is in active litigation. It is one of the patents Neurelis, Inc. is asserting in its Hatch-Waxman (ANDA) campaign against generic Valtoco® (diazepam nasal spray) filers in the District of Delaware, before Judge Maryellen Noreika. The '664 patent is not the subject of any IPR petition that I could find, and I found no Federal Circuit appeal naming it. Details, with confidence levels, below.

⚠️ Source-date caveat up front: several sources I retrieved self-report "last updated" dates after the April 26, 2026 date given in my instructions (e.g., drugpatentwatch "August 25, 2026," TheraRadar data "2026-09-02"). I am reporting those dates literally rather than reconciling them.

1. Cases in which the '664 patent is documented as asserted

A. Neurelis, Inc. v. Padagis LLC et al. — the lead case

This is the strongest '664 hit. Docket Filing 4 ("Report to the Commissioner of Patents and Trademarks for Patent/Trademark Number(s): 11,241,414; 11,793,786; 12,268,664") is PACER-derived and authoritative. The complaint analysis for this docket contains a dedicated "'664 Patent Infringement Allegations" table citing US12268664B1, col. 54:4-6 — i.e., the asserted '664 claim 1 limitations are charted in the complaint. Accused products: Padagis ANDA No. 219320 and Lupin ANDA No. 220394.
Sources: Justia docket 1:25-cv-00821; Ex Parte complaint analysis

B. Neurelis, Inc. v. Strides Pharma Global PTE Limited

drugpatentwatch lists the patents in this case as: 11,241,414; 11,793,786; 12,268,664; 12,324,852; 12,337,061; 12,521,400; 12,599,611. The Strides complaint itself expressly pleads the '664 patent (¶¶22–24) and — importantly — pleads the certificate of correction (see §4 below).
Sources: CourtListener 1:26-cv-00947; drugpatentwatch case page

2. Related Neurelis ANDA cases where the '664 patent is listed but I could not independently confirm it is asserted

These all appear in the same consolidated Delaware family. Aggregator pages (drugpatentwatch, paragraphfour) list '664 among the patents in the consolidated bundle, but the individual dockets do not always name it. Treat these as "listed, not confirmed."

Why I'm hedging on these:

• The '664 patent issued April 8, 2025 — after Padagis I was filed (May 8, 2024). It therefore could not have been in the original Padagis I complaint; any presence is via amended/supplemental pleading, which I did not verify.
• The CourtListener docket for 1:26-cv-00401 reports only 12,521,400 on its "Report to the Commissioner" entry and states: "Date of Expiration of Patent: 03/27/2029. Thirty Month Stay Deadline: 9/27/2026." The drugpatentwatch page for that same docket lists seven patents including '664 — most likely an aggregation of the consolidated family, not the '401 complaint itself.
• The paragraphfour entry for 1:24-cv-00562 lists '664 among "Litigated Patents," while its entry for 1:25-cv-00821 separately lists '12,521,400' — impossible as an original pleading in a July 2025 case, since the '400 patent issued Jan. 13, 2026. These aggregator pages demonstrably merge consolidated dockets, which is why I'm not treating their patent lists as per-case proof.

Sources: CourtListener 1:26-cv-00401; paragraphfour litigation entry; drugpatentwatch 1:24-cv-00562

3. PTAB / post-grant proceedings

No IPR or PGR petition against the '664 patent was found.

• Padagis's three instituted petitions — IPR2025-00464, -00465, -00466 — target the '546, '414, and '786 patents respectively, not '664. Trial was scheduled for June 16, 2026.
• The '664 patent does appear in the PTAB record as Patent Owner Exhibit EX2077 in IPR2025-00464 (Padagis US LLC v. Neurelis, Inc.) (filed Feb. 20, 2026) — i.e., it is evidence in someone else's IPR, not the challenged patent. (Note: one document header renders that case as involving "U.S. PATENT 8,895,546," which I report literally.)
• Earlier, unrelated history: IPR2019-00451, Aquestive Therapeutics, Inc. v. Hale Biopharma Ventures, LLC, invalidated the related '876 patent (affirmed without opinion by the Federal Circuit).

Source: PTAB EX2077

4. ⚠️ Contradiction with the previously generated section — certificate of correction

The earlier summary stated: "A post-issue communication / certificate of correction for US12268664 is recorded as dated January 22, 2026 … I could not retrieve the substance of that correction."

This is now partly resolved and partly contradicted. The Strides complaint (§1.B, ¶22) states the following chronology:

Substance of the correction (new information): the correction was sought "to recite the existence of a Joint Research Agreement, so that it may invoke the protections of 35 U.S.C. § 103(c)(2)." In plain terms, this is a statutory double-patenting / prior-art disqualification fix — Neurelis is trying to disqualify its own joint-research-agreement-derived prior art from being used against the '664 claims. This is a substantive defensive move against the obviousness attack Padagis is running, and it suggests the "about"-range and priority/joint-inventorship issues flagged in the earlier section are live.

I could not retrieve the four corners of the certificate itself, so the corrected claim text remains unverified. The '664 claim 1 language I reported earlier (sourced from the Delaware complaint and drugpatentwatch) should still be checked against the corrected printed claim. Note that the Strides complaint paraphrases '664 as claiming a "pharmaceutical solution for nasal administration consisting of" a benzodiazepine drug, tocopherols/tocotrienols, alcohols, and dodecyl beta-D-maltoside — which is in tension with the earlier summation that '664 uses the open-ended transitional phrase "comprising" (derived from the same complaint family). Flagging this as an unresolved claim-wording discrepancy.

5. Federal Circuit

No CAFC appeal or 2026 Federal Circuit docket naming US 12,268,664 was found. This is consistent with the earlier summary. All identified proceedings remain at the district-court level (plus the collateral IPRs on sibling patents). Absence of a hit is not proof that none exists — the consolidated Delaware case is pre-trial, so there is no final judgment yet to appeal.

6. Summary table — cases directly implicating the '664 patent

Google Patents' own litigation links for this family point to 1:25-cv-00821 and 1:26-cv-00261 in D. Del. (Google Patents). I could not retrieve the Unified Patents portal case list page directly; the Unified Patents links supplied on the Google Patents record match the two dockets above.

7. Explicit uncertainties

1. No IPR/CAFC proceeding on '664 itself — confirmed by search, but absence of evidence is not proof of absence.
2. Per-case patent lists from aggregators are unreliable because they merge consolidated dockets; the only PACER-derived confirmation of '664 as asserted that I obtained is the 1:25-cv-00821 Filing 4 entry and the Strides complaint ¶¶22–24.
3. Corrected claim text of '664 is still unverified; I know only the purpose of the Feb. 10, 2026 certificate (Joint Research Agreement / § 103(c)(2)) and that it was sought Dec. 15, 2025.
4. "Comprising" vs. "consisting of" discrepancy for '664 claim 1, as between the earlier summary and the Strides complaint's paraphrase.
5. No outcome on the merits in any case — all are pre-trial; the earliest reported bench trial date is March 1, 2027.
6. No settlement or license terms are in the public record I retrieved.

I'll start by verifying the structured PTAB data against web sources, since the previously-generated sections flagged a potential IPR that may or may not target this patent.

Let me verify whether any petition targets the '664 patent itself, and pin down the institution/FWD status of the Padagis IPRs.

Let me check whether any petition was ever filed against the '664 patent itself (including a 2026 filing), and nail down the '664 certificate-of-correction date.

Proceedings overview

There are zero AIA trial proceedings on US 12,268,664. The structured "PTAB proceedings on file" block is correct: the USPTO ODP API returns no IPR, PGR, or CBM naming the '664 patent as the challenged patent, and my independent searches (PTAB dockets via DocketAlarm, ipverse/GreyB, DrugPatentWatch's PTAB tables, the P-TACTS petition repository, and the Padagis paper trail) found none. Breakdown, therefore, is 0 active / 0 invalidated / 0 sustained / 0 settled / 0 denials — a null result, not a score.

The bottom line for a defendant is not "the patent survived and is hardened." It is the opposite of the usual signal, and it has two edges:

1. The '664 patent is untested at the PTAB. It issued 2025-04-08 and was only asserted in D. Del. on 2025-07-02, well after Padagis's 2025-01-17 IPR wave was aimed at its three older siblings (U.S. 8,895,546; 11,241,414; 11,793,786). No one has run — or apparently can still run — an AIA petition against it from the defendant side.
2. But the family's PTAB record is catastrophic for the patent owner, and Neurelis has separately armed the '664 patent with a Joint Research Agreement ("JRA") certificate of correction while the Board denied it the same correction on '546, '414 and '786. That single asymmetry is the most important fact in this report.

So the honest defensive read is: there is no PTAB win to cite against this patent; you cannot rely on claim cancellation; but the sibling litigation gives you an established Gwozdz-over-Meezan invalidity theory, and the '664 certificate of correction is the patent owner's counter-weapon, which you must plan around.

Proceedings naming US 12,268,664 as the challenged patent

None. For completeness, the only appearances of the '664 patent in any PTAB record I located are as evidence, not as the subject patent:

• IPR2025-00464, Ex. 2077 — "U.S. Patent No. 12,268,664" (filed 2026-02-20 by Patent Owner).
• IPR2025-00464, Ex. 2078 — "664 Certificate of Correction" (filed 2026-02-20 by Patent Owner).

Contradiction resolved from the previously-generated section. The earlier summary flagged that "one document header renders the case as involving 'U.S. PATENT 8,895,546' while the exhibit itself is the '664 patent," and treated it as unresolved. It is now explained: IPR2025-00464 is captioned U.S. Patent No. 8,895,546; Neurelis filed the '664 patent and its certificate of correction as exhibits in that '546 proceeding to argue that a correction the Office granted on the '664 patent proves the same correction is proper on '546. The '664 patent is not under review in that case.

Adjacent proceedings you must know about (NOT proceedings on the '664 patent)

These are listed because they supply the prior art, the estoppel map, and the credibility of any invalidity theory you would raise against '664.

IPR2019-00451 — Aquestive Therapeutics, Inc. v. Neurelis, Inc.

• Type: Inter Partes Review
• Subject patent: U.S. 9,763,876 (a family sibling; not the '664 patent)
• Filed: 2019-01-29
• Status: Final Written Decision issued, all claims held unpatentable; affirmed on appeal.
• Judge panel: Not retrieved in my searches — I will not guess names.
• Petition grounds: § 103 obviousness, all 36 claims of the '876 patent, over Gwozdz (WO 2009/120933) in view of Meezan (US 2006/0046962).
• Final Written Decision: 2020-08-06 — all claims 1–36 of the '876 patent held unpatentable as obvious. The panel also determined the challenged claims were not entitled to the priority date of provisional application 61/040,558 (later confirmed and relied on by the Acting Director in the 2025 Padagis cases).
• Appeal: Neurelis appealed to the Federal Circuit (docket 21-1038); Aquestive was appellee. The Federal Circuit affirmed under Fed. Cir. R. 36 (reflected in IPR2025-00464 Ex. 1068 and Ex. 2028).
• Settlement: None — the case ran to judgment and affirmance.
• Defensive value: This is the family's load-bearing invalidity precedent. Everything Padagis later did flows from it. Any '664 invalidity theory rooted in Gwozdz-over-Meezan has already been litigated to judgment on near-identical claims — but see the certificate-of-correction problem below.

IPR2025-00464 — Padagis US LLC v. Neurelis, Inc.

• Type: Inter Partes Review
• Subject patent: U.S. 8,895,546 ('546) — not the '664 patent
• Filed: 2025-01-17 (accorded filing date 2025-01-17). RPIs: Padagis LLC, Padagis US LLC, Padagis Israel Pharmaceuticals Ltd.
• Status: Trial Instituted (institution granted 2025-09-16).
• Judge panel: Grace Karaffa Obermann, Annette R. Reimers, and Timothy G. Majors (Reimers authored the institution decision).
• Petition grounds: § 103 — claims 1–22 obvious over Gwozdz PCT (WO 2009/120933) in view of Meezan (US 2006/0046962). Petitioner also pressed collateral estoppel off the '876 FWD on priority, Gwozdz's prior-art status, and the Gwozdz+Meezan combination.
• Institution decision: 2025-09-16 — instituted on all challenged claims, citing SAS Inst. Inc. v. Iancu and 37 C.F.R. § 42.108(a). (institution decision PDF)
• Threshold discretionary fight (important, and designated informative): Neurelis sought discretionary denial under Fintiv, arguing the D. Del. trial (2026-02-09) would precede the FWD due date (2026-09-17). Acting Director Coke Morgan Stewart denied the request on 2025-07-16 and referred the petitions to the Board, reasoning that the examiner's contrary priority determination "directly contradicts" the '876 FWD and "raises concerns of material error such that review ... is an appropriate use of Office resources" (Advanced Bionics). Director John Squires designated the decision informative on 2026-01-09. (Acting Director decision)
• Certificate-of-correction motion (directly relevant to '664): Neurelis moved (2025-12-23) for leave to petition the Director to add a JRA disclosure so it could invoke the pre-AIA 35 U.S.C. § 103(c)(2)(C) safe harbor and eliminate Gwozdz as prior art. The Board denied leave on 2026-02-06, finding both an insufficient basis and "substantial prejudice to Petitioner," and denied rehearing on 2026-03-20. The rehearing decision is where the '664 patent enters the record. (rehearing decision)
• Other procedural events: Patent Owner did not file a motion to amend; the parties stipulated to revised Due Dates 2 and 3 (Due Date 8 / oral argument set for 2026-06-16); a statutory disclaimer on the '546 patent was filed 2026-01-02 (Ex. 2075) — the disclaimed claims are not stated in the materials I retrieved.
• Final Written Decision: Not issued in the records I retrieved, and statutorily due 2026-09-17 (one year from institution). I found no FWD text. This is a live, unresolved case as of today.
• Settlement / termination: None. The case remains on the trial track.
• Appeal: None yet (no FWD to appeal).
• Defensive value: If the Board invalidates the '546 claims on Gwozdz-over-Meezan on 2026-09-17, it is strong persuasive — not preclusive — ammunition against the '664 claims, which recite the same solvent system plus a dodecyl β-D-maltoside enhancer. Conversely, an FWD sustaining the claims would sharply weaken that line.

IPR2025-00465 — Padagis US LLC v. Neurelis, Inc.

• Type: Inter Partes Review
• Subject patent: U.S. 11,241,414 ('414) — not the '664 patent
• Filed: 2025-01-17
• Status: Trial Instituted (institution granted 2025-09-16).
• Judge panel: Same three-APJ panel (Obermann, Reimers, Majors); the 2026-03-20 rehearing decision is captioned to all three proceedings.
• Petition grounds: § 103 — claims 1–18 obvious over Gwozdz (WO 2009/120933) in view of Meezan (US 2006/0046962); Petitioner also advanced § 112 written-description arguments on priority and a vitamin E construction (per the patent's col. 30 definition). Petitioner asserted collateral estoppel off IPR2019-00451.
• Institution decision: 2025-09-16 — instituted. Note the Acting Director's decision described this as Paper 12 (2025-07-16) and cited it as informative.
• Final Written Decision: Not issued in the records retrieved; due 2026-09-17.
• Settlement / termination: None identified.
• Appeal: None yet.

IPR2025-00466 — Padagis US LLC v. Neurelis, Inc.

• Type: Inter Partes Review
• Subject patent: U.S. 11,793,786 ('786) — not the '664 patent. (Respondent application number 17/228,514 — the '664 patent's own parent application, per the family chain.)
• Filed: 2025-01-17
• Status: Trial Instituted (institution granted 2025-09-16).
• Judge panel: Same panel; extensive motion practice (motions to strike, motions to exclude, pro hac vice).
• Petition grounds: § 103 over Gwozdz in view of Meezan, plus priority/written-description attacks and collateral estoppel.
• Institution decision: 2025-09-16.
• Trial-stage milestones (public docket): Patent Owner Response (public redacted) 2026-01-06; Petitioner Reply 2026-03-31; Patent Owner Sur-reply 2026-05-12; Board order setting oral argument 2026-05-21; oral argument papers filed 2026-06-11. (case docket)
• Final Written Decision: Not issued in the records retrieved; the FWD due date is 2026-09-17 (shared across the three Padagis IPRs).
• Settlement / termination: None identified.
• Appeal: None yet.

Strategic summary

Claim status of the '664 patent: 100% UNTESTED. No claim of US 12,268,664 has been canceled, confirmed, or construed by the Board. The three proceedings where Gwozdz is being litigated concern the '546, '414 and '786 patents. Any statement that "'664 claims 1–X have been canceled" would be false, and I found no such statement in any source. For the family, by contrast, the picture is stark: claims 1–36 of the '876 patent were canceled (FWD 2020-08-06, affirmed Fed. Cir. 21-1038), and the '546, '414 and '786 claims are under instituted review with FWDs due 2026-09-17.

Estoppel landscape — the key practical point. Because no IPR was filed against the '664 patent, § 315(e)(2) IPR estoppel does not attach to it. Padagis (and its RPIs) are estopped only as to grounds they raised or reasonably could have raised against the '546, '414 and '786 patents. They are free to run any prior-art ground against '664 in D. Del. The limiting doctrine on '664 is instead § 315(b): the one-year clock runs from service of a complaint alleging infringement of that patent. Neurelis asserted the '664 patent in D. Del. on 2025-07-02 (Neurelis, Inc. v. Padagis LLC et al., 1:25-cv-00821); service followed, which means the § 315(b) window for Padagis and Lupin to petition against '664 appears to have closed in mid‑2026 — and no petition was filed. I flag this as an inference from the docket dates, not a legal conclusion: confirm actual service dates before relying on it. The corollary is the opportunity: any party that has not been served with a complaint on the '664 patent (another ANDA filer, or a defensive aggregator such as Unified Patents) remains unbarred and can still petition. I found no Unified Patents involvement in this family.

Pattern signals. The same petitioner (Padagis, with its RPIs) filed three coordinated IPRs on this family on the same day (2025-01-17), all on a single theory (Gwozdz + Meezan), and is pressing them through oral argument. Neurelis has litigated PTAB threshold issues aggressively: a Fintiv discretionary-denial request (denied, and designated informative against it), a certificate-of-correction motion (denied and rehearing denied), a pending statutory disclaimer on the '546 patent, and — critically — it took the ex parte route to a certificate of correction on the '664 patent specifically. That last move is a "hedge": the patent owner appears to be fortifying its newest, still-unreviewed family member with the JRA safe harbor that the Board refused it on the three patents actually under attack. Practical warning: if you assert a Gwozdz-based § 103 ground against the '664 patent, expect Neurelis to invoke § 103(c)(2)(C) via the corrected '664 specification — a defense its three siblings do not have.

Recommended next steps

• Do not represent that any claim of the '664 patent has been invalidated. Nothing has been. Your defensive value on this patent comes from (a) the siblings' litigation, and (b) the § 315(b)/§ 315(e) posture, not from a cancellation.
• Docket the FWD date: 2026-09-17. Three FWDs (IPR2025-00464/-00465/-00466) are statutorily due that day. They will not state anything about the '664 claims, but a holding of unpatentability on '546/'414/'786 claims on the Gwozdz-over-Meezan theory materially changes the settlement and invalidity calculus for '664. Check PTAB E2E (https://ptacts.uspto.gov/ptacts/) on and after that date.
• Get the '664 certificate of correction itself. Per the description of IPR2025-00464 Ex. 2078, the request (filed 2025-12-15) stated that "[t]he Certificate of Correction merely adds the names of the parties to a joint research agreement and, therefore, does not constitute new matter pursuant to MPEP § 2156." If that description is accurate, the '664 claims were not amended — which resolves the previously-generated section's warning that the correction "could have changed claim wording." I am relying on document snippets, so obtain the printed certificate from the USPTO before relying on it.

Contradiction flagged (do not paper over): The previously-generated section states the '664 certificate of correction is "recorded as dated January 22, 2026" (GreyB/Pharsight), whereas the Board's 2026-03-20 rehearing decision refers to "the February 10, 2026, grant of the certificate of correction on the related patent" (Ex. 2078). These cannot both be the grant date. The request was filed 2025-12-15. I could not resolve which date is the operative grant date; verify directly against USPTO Patent Center / the printed certificate.

• If you are Padagis or Lupin and want an AIA route against '664, assume your § 315(b) window has closed and pursue validity in D. Del. instead. If you are a different ANDA filer not yet served, the window is open: the Gwozdz-over-Meezan ground is fully briefed, has a Federal Circuit affirmance on the '876 patent, and has already produced institution on all three challenged siblings — but budget for the '664 JRA safe-harbor fight.
• Watch the '546 statutory disclaimer (filed 2026-01-02, Ex. 2075). If it disclaims claims implicated in the '546 IPR, it may affect the scope of the surviving '546 claims and, indirectly, how Neurelis positions the '664 patent in the Delaware cases. The disclaimed claim numbers are not stated in the materials I retrieved.

Sources for the opinions cited above: IPR2025-00464 institution decision (2025-09-16) and Acting Director referral decision (2025-07-16, Paper 12, informative) via the links above; IPR2025-00464 rehearing decision (Paper 39, 2026-03-20); IPR2025-00466 docket via ipverse/GreyB; Aquestive's counsel announcement of the '876 FWD (https://hbiplaw.com/library-posts/announcements/hoffmann-baron-on-behalf-of-aquestive-inc-successfully-invalidates-all-claims-of-orange-book-patent-for-valtoco-nasal-spray-in-aquestive-therapeutics-inc-v-neurelis-inc/); and the D. Del. complaint in 1:26-cv-00401 (https://www.courtlistener.com/docket/73167119/1/neurelis-inc-v-padagis-llc/).

If you need the IPR2019-00451 panel composition or the full FWD text, those are public at the PTAB's E2E/PTACTS site under case IPR2019-00451 — I did not retrieve the panel names and have not guessed them.

I'll research the assignment chain for US 12,268,664. Let me run several targeted searches.

Useful leads. Let me dig into the ownership chain and the parties.

Let me hunt for the specific recorded assignment records and any security interests.

I hit the session's tool-call limit before I could pull the Assignment Center record itself. Below is the reconstruction from the records I did retrieve, with the gaps flagged rather than papered over.

Ownership chain — US 12,268,664 ("Administration of Benzodiazepine Compositions")

Critical sourcing caveat up front: I could not open assignmentcenter.uspto.gov / assignment.uspto.gov in this session. The USPTO Assignment Center is a JS-driven database that is not surfaced through general web indexing, and no retrieved page reproduced a reel/frame number or a recorded correspondent for this patent. What follows is built from (a) the Google Patents "reassignment" legal-event block on the patent record you supplied, (b) the PTAB EX2078 certificate-of-correction packet (which reproduces Patent Center filing metadata), and (c) the EX2047 Hale declaration in IPR2025-00464. Anything I could not verify is marked NOT RETRIEVED rather than estimated. Do not treat the reel/frame slots below as filled.

Inventors

Seven named inventors, and the notable feature is that they were not employed by one company — this is a genuine three-party joint-research invention:

Unusual patterns — what is and is not present:

• Present and material: no common employer. Inventors span Questcor, Particle Sciences, and Aegis, with HBV supplying two (Hale, and Maggio as consultant). This is why the patent needed a post-issuance Joint Research Agreement statement (see timeline) naming Particle Sciences, Inc. and Questcor Pharmaceuticals, Inc. (as predecessor-in-interest to Hale BioPharma Ventures, LLC). The JRA statement is a defensive priority/prior-art move — it is the mechanism (pre-AIA § 103(c) / AIA § 102(b)(2)(C)) for attempting to disqualify the Gwozdz PCT reference, which shares a named inventor (Gwozdz) with this patent. Note the IPR history: the Board in the earlier '876 IPR (IPR2019-00451) found the '558 Provisional did not support the "alkyl glycoside" limitation, which knocked out the March 28, 2008 priority date for those claims and made Gwozdz PCT prior art. The JRA correction reads as an attempt to close that gap for the '664 patent.
• Not present: mass inventor departure. I found no evidence of all inventors leaving one assignee within 12 months. There was no single original assignee to leave — the asset moved (Questcor → HBV), and two inventors (Hale, Maggio) were on the acquiring side from the outset.

Original assignee

On the face of the issued patent: Neurelis, Inc., San Diego, CA — Google Patents lists Neurelis as both Original Assignee and Current Assignee for US 12,268,664.

Functionally, the originating owner of the priority filing: Hale BioPharma Ventures, LLC (Encinitas, CA), expressly as successor-in-interest to Questcor Pharmaceuticals, Inc. HBV filed the non-provisional application 12/413,439 on March 27, 2009 after executing an Asset Purchase Agreement dated June 27, 2008 acquiring the QSC-002 program from Questcor (EX2047 ¶¶9–11, ¶16–17). The earlier family members bear HBV as assignee on their faces (e.g., US 8,895,546 and US 9,763,876 list "Hale Biopharma Ventures, LLC"); later members list Neurelis.

Does the assignee ship a product embodying the claims? Yes — unambiguously.

• VALTOCO® (diazepam nasal spray), 5 / 7.5 / 10 mg, NDA 211635, approved January 10, 2020. US 12,268,664 is Orange Book–listed against it.
• Claims 1–18 are directed to a diazepam nasal solution (1–20 mg diazepam in a tocopherol/ethanol system with n-dodecyl β-D-maltoside), which reads directly on the commercial product.

Primary line of business: commercial-stage, privately held specialty pharmaceutical company (CNS/neurology rescue therapy). Not a licensing vehicle: it funds development, holds the NDA, and handles commercialization.

Current status: Operating. VC-backed and capitalized — Hale BioPharma Ventures participated in a ~$114.76M Series D on March 8, 2021, alongside earlier rounds. No bankruptcy, no Chapter 7/11, no dissolution found. (Separately, HBV itself is an operating venture-builder founded 2006 with ~22 investments — a real fund, not a paper entity.)

Assignment timeline

Two title transfers are visible on the record, both carrying a recording date of 2024-07-24. Execution dates and all reel/frame numbers: NOT RETRIEVED.

• Execution date NOT RETRIEVED / recorded 2024-07-24 — Reel NOT RETRIEVED/NOT RETRIEVED

  • Conveyance: Assignment (as reflected in Google Patents "reassignment" event)
  • Assignor: CARTT, STEVE; MEDEIROS, DAVID; GWOZDZ, GARRY THOMAS; LOXLEY, ANDREW; MITCHNICK, MARK; HALE, DAVID F.; MAGGIO, EDWARD T. (all seven named inventors)
  • Assignee: HALE BIOPHARMA VENTURES, LLC
  • Correspondent: NOT RETRIEVED. (Post-issuance correspondence of record for this property is DLA Piper LLP (US), c/o Dennis J. Parad, Reg. No. 60,280, and prior prosecution correspondence was Wilson Sonsini per EX2047 ¶17. Neither is confirmed as the assignment-recording correspondent — do not conflate.)
  • Context: origin-of-title — inventors (across Questcor / Particle Sciences / Aegis) assigning to the HBV vehicle that acquired the QSC-002 program. The late 2024 recording date coincides with the filing of continuation 18/782,666 and is consistent with re-recording the pre-existing chain against the new application/serial number, not with a fresh 2024 acquisition.

• Execution date NOT RETRIEVED / recorded 2024-07-24 — Reel NOT RETRIEVED/NOT RETRIEVED

  • Conveyance: Assignment (internal transfer within the same sponsorship group)
  • Assignor: HALE BIOPHARMA VENTURES, LLC
  • Assignee: NEURELIS, INC.
  • Correspondent: NOT RETRIEVED.
  • Context: internal reorg / affiliate transfer — the same asset moving from the Hale venture vehicle to the operating company that holds the NDA and commercializes Valtoco. Again recorded contemporaneously with the 2024 continuation filing.

Adjacent title-relevant records (not assignments, but part of the ownership story)

• 2025-04-08 — Patent granted to Neurelis, Inc.; "Assigned to NEURELIS, INC." event posted.
• 2025-12-15 — Request for Certificate of Correction filed under 37 C.F.R. § 1.323 (applicant's mistake), attorney docket 148570-014311, signed /Dennis J. Parad/, Reg. No. 60,280, DLA Piper LLP (US), 1650 Market Street, Suite 5000, Philadelphia, PA 19103; Customer No. 185628; Deposit Account 60-4406; $172.00 fee.
• 2026-02-10 — Certificate of Correction issued (signed John A. Squires, Director). It inserts, immediately after paragraph [001], a "JOINT RESEARCH AGREEMENT" statement: "The subject matter disclosed was developed and the claimed invention was made by or on behalf of the following parties to a joint research agreement: Particle Sciences, Inc. and Questcor Pharmaceuticals, Inc. (as predecessor-in-interest to Hale BioPharma Ventures, LLC)." This is a specification-only addition; per the request it "does not constitute new matter." The claims themselves were not amended by the correction — which is useful, because it means your previously generated claim analysis (claim 1 as recited in the D. Del. complaints) stands on the corrected patent.

⚠️ Contradiction to flag (inherited from the prior section, now partially resolved): the prior summary recorded the certificate of correction as dated January 22, 2026 with unknown substance. The EX2078 packet shows the correction was signed/sealed February 10, 2026, with an electronic notification dated 02/10/2026; the "January 22, 2026" figure appears to be a database posting/processing date rather than the certification date. Substance is now known and stated above. Also flag as before: the EX2078 header renders the caption as "U.S. PATENT 8,895,546" while the exhibit contents are the '664 patent's correction papers — reporting literally, not reconciling.

What I could not retrieve (material gaps)

1. Reel/frame numbers for either assignment.
2. Execution dates for either assignment.
3. Recorded correspondents (the attorney/firm that filed the recordings) — the single most diagnostic field for this task.
4. Whether there are additional recorded instruments at the application level (e.g., a 2009-instrument recorded against 12/413,439, or any security agreement / lien from Neurelis's venture-debt or equity financings). I found no evidence of a recorded security interest, but absence of evidence here is weak, given I could not query the database directly.

Verification: query https://assignmentcenter.uspto.gov/ (or https://assignment.uspto.gov/patent/index.html) by patent number 12268664, and independently by application 18/782,666 and parent 12/413,439, to capture the reel/frame and correspondent fields.

Timeline diagram

timeline
    title Ownership of US 12268664
    2006 : Hale BioPharma Ventures formed
    2008 : Questcor files provisional 61/040/558
         : HBV buys QSC-002 from Questcor
    2009 : HBV files non-provisional 12/413/439
    2024 : Inventors assign to HBV recorded
         : HBV assigns to Neurelis recorded
    2025 : Patent issues to Neurelis
         : Neurelis sues Lupin and Padagis
    2026 : Certificate of correction issued

(Note: to keep the parser safe I stripped slashes from the application/provisional numbers; if your renderer tolerates them, 61/040,558 and 12/413,439 may be restored — commas are the risky token.)

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. There are two entities in the chain, and neither is a licensing-only shell. Hale BioPharma Ventures, LLC carries the "Ventures" suffix but is a real venture-builder (founded June 2006, ~22 documented investments, offices at 1042-B North El Camino Real, Encinitas, CA, a normal commercial address rather than a registered-agent drop). Its assignee role was asset acquisition, not assertion. Neurelis, Inc. is a commercial-stage pharma with an approved NDA. No single-member Delaware/Texas LLC, no registered-agent address, no assertion activity, on either link.

2. Known asserter in the chain — NOT PRESENT. Neurelis, Inc. and Hale BioPharma Ventures, LLC appear on none of the enumerated asserter lists (Acacia, Marathon, IV, IPNav, Wi-LAN, Mosaid/Conversant, Vringo, Pendrell, Innovatio, MPHJ, Lumen View, Round Rock, Document Generation Corp, Spangenberg entities), and I surfaced no Unified Patents or RPX high-frequency-plaintiff listing for either. The litigation in which Neurelis appears — IPR2025-00464 (Padagis), the D. Del. actions 1:25-cv-00821, 1:26-cv-00261, 1:26-cv-00401, and the earlier IPR2019-00451 (Aquestive, then styled Aquestive v. Hale BioPharma Ventures) — is Neurelis as patent owner defending an approved product, i.e., the inverse posture.

3. Repeat correspondent across the chain — UNCLEAR / NOT PRESENT as an NPE tell. The only correspondents I could identify are DLA Piper LLP (US) — Dennis J. Parad, Reg. No. 60,280, Customer No. 185628 (certificate-of-correction packet, docket 148570-014311) and Wilson Sonsini (original 2009 filing, EX2047 ¶17). DLA Piper recurs across Neurelis's Valtoco patent family (the 148570-xxxxxx docket series), but that recurrence reflects one operating-company client using one firm, which is the opposite of the shell-LLC pattern this signal is designed to catch. Critically, I could not retrieve the correspondents on the two assignment recordings — so this signal cannot be fully cleared. Re-check it directly.

4. Cascading transfers — NOT PRESENT. Two recorded instruments, both within one sponsorship group, both recorded 2024-07-24. No chain of unrelated LLCs, no <24-month cascade, no shared-correspondent-address pattern observable.

5. Pre-litigation transfer — NOT PRESENT. The HBV → Neurelis transfer recorded 2024-07-24 precedes the first named suit (1:25-cv-00821), but (a) Neurelis had already been the named assignee on earlier family members (11,793,786, 11,241,414, 12,324,852, 12,337,061), so it has owned the family for years, and (b) the recording is synchronised with the continuation filing, not with the complaint. This is docket hygiene, not standing engineering.

6. Bankruptcy fire-sale — NOT PRESENT. No Chapter 7/11 for Neurelis or HBV found. The 2008 Questcor → HBV transfer was a voluntary carve-out of a non-core development-stage program (Project QSC-002, then without a lead candidate, clinical data, or regulatory testing — EX2047 ¶5), not a distressed sale. Questcor was later acquired by Mallinckrodt (2014) as a going concern.

7. Privateering — NOT PRESENT. Neurelis is asserting its own Orange Book patents against actual generic ANDA filers (Padagis ANDA 219320, Lupin ANDA 220394) for the product it sells. No operating company is using Neurelis as a proxy to attack competitors.

8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. No neutralisation.

Verdict

Operating-company assertion.

The chain runs inventors (Questcor + Particle Sciences + Aegis) → Hale BioPharma Ventures, LLC → Neurelis, Inc. per the two Google Patents reassignment events recorded 2024-07-24 (reel/frame NOT RETRIEVED), terminating at a commercial-stage specialty pharma that holds NDA 211635 for VALTOCO and whose claims read on that product. Zero of the eight NPE signals is present; the only unresolved item is signal 3, and that is an evidence gap (recorded correspondents not retrievable), not a hint of shell-LLC recurrence. The litigation posture — Neurelis defending its own approved-product patents in IPR2025-00464 and asserting them in 1:25-cv-00821 / 1:26-cv-00261 / 1:26-cv-00401 against Padagis and Lupin — is the classic operating-company Hatch-Waxman posture. The 2026 certificate of correction, which adds a Joint Research Agreement statement for Particle Sciences and Questcor, is a defensive prior-art move (aimed at the Gwozdz PCT), not a title change.

Verify at: https://assignmentcenter.uspto.gov/ — search patent 12268664, then application 18/782,666 and parent 12/413,439, to fill the reel/frame and correspondent fields that this session could not retrieve.

I'll search for the authoritative citation records for US 12,268,664.

I found a USPTO-hosted copy with the "(56) References Cited" section. Let me retrieve the full citation list.

Good progress. Let me retrieve the remaining portions of the front-page reference list (second column, foreign documents, and other publications).

Prior Art Cited on US 12,268,664 B1 ("the '664 patent")

Scope note. I retrieved the '(56) References Cited' section as it appears on the face of the printed patent (via the Patent Owner's EX2077 copy filed in Padagis US LLC v. Neurelis, Inc., IPR2025-00464, and the corresponding front-page image text). The retrieved list is partial — the OCP/OCR transcriptions I could reach truncate in two places (a U.S.-patent gap between 5,556,940 and 6,090,925, and the entire 'Other Publications' section). I flag every gap rather than fill it by inference. Where a reference's subject matter is confirmed by the specification text supplied to me, I mark it [verified from spec]. Where I am describing content from my own background knowledge, I mark it [unverified] and give a confidence level. I did not fabricate any description.

Also carried forward from the prior section: a Certificate of Correction dated 2026-01-22 is of record for this patent, and its substance remains unretrieved. Every §102 statement below is therefore keyed to the claim text as it appears in the D. Del. complaint and drugpatentwatch, not to the corrected printed claim.

A. U.S. Patent Documents (front page, col. 1)

B. U.S. Patent Application Publications (front page, col. 2)

C. Foreign Patent Documents (front page)

D. Other Publications

Not retrieved. The 'OTHER PUBLICATIONS' block on the '664 front page is followed by "(Continued)" and the text was truncated in every accessible copy. This is the single largest evidentiary hole in this analysis, because for a formulation patent of this type the examiner's non-patent-literature citations (journal articles on intranasal benzodiazepine pharmacokinetics, dodecyl maltoside absorption enhancement, and Vitamin E-based non-aqueous vehicles) are frequently more probative than the patent citations. I cannot report this section and am not going to guess at its contents.

E. §102 analysis

E.1 Threshold problem: the cited art does not map onto the claim

Claim 1 is a single, highly specific combination:

1–20 mg diazepam dissolved in 45–85% w/w tocopherol/tocotrienol (genus including tocophersolan) + 25–40% w/w of a C2–C5 alcohol/benzyl alcohol + 0.01–1% w/v dodecyl β-D-maltoside, wherein the composition is a solution and contains less than 1% water, in a nasal spray formulation.

Anticipation under §102 requires all elements in a single reference, arranged as claimed. On the record I retrieved:

• None of the 12 benzodiazepine-chemistry patents (items 1–9, 11, 13, 16 in Table A) discloses anything about a Vitamin E/alcohol vehicle, a maltoside enhancer, a water limit, or a nasal spray. They are active-ingredient/manufacture references. Anticipation of claims 1–18: No.
• The nanoparticulate/microparticulate patents (Liversidge 5,145,684; Fessi 5,118,528; Violante 4,997,454; Kipp 6,607,784; and the Mathiowitz/Woiszwillo family) go to the suspension embodiments the specification describes — but claim 1 affirmatively requires a solution and is substantially free of microparticles/nanoparticles. Those references are therefore directed away from claim 1 as granted (though they remain relevant to §112 written-description/support questions about the suspension disclosure). Anticipation of claim 1: No.
• US 6,193,985 (Sonne, Vitamin E TPGS) is the only reference I can affirmatively tie to a claim element: claim 1's tocopherol genus expressly includes tocophersolan (= Vitamin E TPGS). But 6,193,985 is a TPGS-chemistry/utility reference; it does not disclose diazepam dissolved in TPGS plus an alcohol plus dodecyl β-D-maltoside under 1% water as a nasal spray. Anticipation of claim 1: No — but note the internal tension flagged below.

E.2 The genuine §102 candidates (ranked, with confidence)

1. US 2006/0147386 A1 (Wermeling, pub. 7/2006) — the only front-page reference that on its face targets intranasal benzodiazepine administration. It is §102(b) art. If it discloses a diazepam/benzodiazepine nasal formulation, it is the closest single-reference challenge. I could not retrieve its text; verify before relying. Confidence in relevance: moderate-high.
2. US 2006/0046969 A1, 2007/0298010 A1, 2008/0194461 A1, 2008/0200418 A1 (Maggio) — the alkyl-glycoside/Intravail family (dodecyl maltoside is the hallmark Intravail surfactant). If any of these discloses an alkyl-maltoside-enhanced transmucosal formulation with a benzodiazepine, it directly reads on the dodecyl β-D-maltoside limitation. The two 2008 publications are §102(e)-only (filing-date dependent). Confidence in relevance: moderate.
3. US 2008/0070904 A1 and 2008/0076761 A1 (Jamieson et al., pub. March 2008) — published essentially contemporaneously with the 2008-03-28 priority date. Whether these are §102(a)/(e) art to this continuation (which claims the 2008-03-28 benefit date) turns on exact publication and filing dates that I could not verify. Timing-critical; verify.
4. US 6,551,578 B2 (Adjei) and US 5,182,258 (Chiou) — likely transmucosal/nasal delivery platform patents; unverified content.

Bottom line on §102: on the retrieved record, no cited reference anticipates any of claims 1–18 individually. The cited set is a §103-style mosaic (benzo actives + delivery vehicles + enhancers), and the real anticipation/obviousness fight over this patent is happening in the IPRs, not in the examiner's citation list.

E.3 Two items I want to flag rather than smooth over

• Internal tension between claim 1 and the specification. Claim 1's tocopherol Markush group recites tocophersolan (Vitamin E TPGS), yet the specification states that for solution embodiments "the tocopherol is substantially or completely free of Vitamin E TPGS" and that "the compositions described herein exclude Vitamin E TPGS." A reference disclosing TPGS-based diazepam solutions is squarely within the literal claim but outside the described invention. That mismatch is a §112 and claim-construction vulnerability worth noting, and it is exactly the sort of thing a §102/§103 petitioner would exploit with 6,193,985 or a TPGS formulation reference.
• Contradiction carried from the prior section (still unresolved). The IPR2025-00464 document header renders the proceeding as involving "U.S. PATENT 8,895,546" while the exhibit itself is the '664 patent. I report that literally; I have not reconciled it. Combined with the unretrieved 2026-01-22 Certificate of Correction, the claim text used above should be treated as provisional pending retrieval of the corrected printed claims.

Explicit uncertainties in this section

1. The U.S. patent citation list is incomplete — at minimum a gap between 5,556,940 and 6,090,925, which plausibly contains US 5,661,130 (the alkyl-glycoside patent the specification itself cites) and other mid-1990s art.
2. The 'Other Publications' section was not retrievable at all.
3. EP 0 396 777 A1's publication date was lost in OCR.
4. Subject-matter descriptions marked [unverified] are my own recollections, not retrievals. They should be independently confirmed before being used in any filing.
5. No CAFC activity was found for the '664 patent (consistent with the prior section).

I'll research the prior art references on the '664 patent's face and the art asserted against the Neurelis family in the related IPRs, since the provided Google Patents text omits the "References Cited" section.

I have strong grounding on the family's prior-art dispute. Let me pin down three specifics: what "Cartt '784" is, the Meezan reference, and the '664 certificate-of-correction / priority posture.

Obviousness Analysis — U.S. Patent 12,268,664 (“the ’664 patent”)

Scope note before the analysis. The Google Patents full text supplied for US 12,268,664 does not contain a “References Cited” section or a claims section; a literal “Prior Art” tab was not available to me. I therefore reconstructed the prior art from two grounded sources: (i) the ’664 front-page citations I was able to retrieve (USPN Office capture of the Cartt et al. face page, showing the (56) list beginning at U.S. Pat. Nos. 3,102,116 … 5,268,461 …), and (ii) the references that have actually been applied against this exact patent family in IPR2019-00451 (Aquestive v. Hale Biopharma, ’876 patent) and IPR2025-00464/-00465/-00466 (Padagis v. Neurelis, ’546/’414/’786 patents). Where I could not verify something, I say so.

I build on, and do not repeat, the patent summary and litigation summary already generated. Two cross-references are load-bearing here and are flagged again below: the certificate of correction purpose (Joint Research Agreement / 35 U.S.C. § 103(c)(2)) and the “comprising” vs. “consisting of” discrepancy.

1. The claim to be analyzed

Claim 1, as charted verbatim in the D. Del. complaint (Neurelis v. Padagis/Lupin, 1:25-cv-00821) at US12268664B1, col. 54:4–25:

“1. A pharmaceutical composition for intranasal administration comprising:
– about 1 to about 20 mg of diazepam dissolved in
– about 45% to about 85% (w/w) of one or more natural or synthetic tocopherols or tocotrienols selected from the group consisting of: α-tocopherol, β-tocopherol, γ-tocopherol, δ-tocopherol, α-tocotrienol, β-tocotrienol, γ-tocotrienol, δ-tocotrienol, tocophersolan, any esters thereof and any combinations thereof, and
– about 25% to about 40% (w/w) of one or more alcohols selected from the group consisting of ethanol, propyl alcohol, butyl alcohol, pentanol, benzyl alcohol, and any combinations thereof, and
– about 0.01% to about 1% (w/v) of dodecyl β-D-maltoside,
wherein the pharmaceutical composition is a solution and contains less than 1% water, wherein the pharmaceutical composition is in a pharmaceutically-acceptable spray formulation.”

Claims 2–18 are dependent (α-tocopherol; 10–250 and 20–50 mg/mL; 2–10 mg; 60–75% tocopherol; added excipients; 25–250 / 50–150 / ~100 µL fill volumes; water <0.5% / <0.25% / <0.1% / none; an impurity limit of not more than 0.01% diazepam impurity at three ICH stability conditions; and spray-content-uniformity characteristics). Claim 18 and the corrected claim text remain unverified (see §9).

2. Governing framework

Because the ’664 patent is a continuation claiming priority to a March 28, 2008 provisional (61/040,558) through a March 27, 2009 parent (12/413,439), the claims are presumptively examined under pre-AIA 35 U.S.C. § 103(a) — the analysis the parties are actually running (Neurelis invoked pre-AIA § 103(c)(2), not the AIA analogue, § 102(b)(2)(C)). Framework:

• Graham v. John Deere factors; KSR Int’l v. Teleflex (combination of known elements, “design incentives,” “finite number of identified, predictable solutions”).
• Overlapping ranges: In re Peterson, 315 F.3d 1325 (Fed. Cir. 2003); In re Woodruff, 919 F.2d 1575 (Fed. Cir. 1990); routine optimization of ranges: In re Aller, 220 F.2d 454 (CCPA 1955).
• Optimizing a disclosed parameter or a known composition: In re Boesch, 617 F.2d 272 (CCPA 1980).
• Where the claim is a composition defined by a newly discovered property, the property does not confer patentability: the “new property of a known composition” line.

3. Person of ordinary skill in the art (POSITA)

On this record (the parties’ competing definitions in the Padagis IPRs; the Donovan declaration, EX1062 in IPR2025-00464), the POSITA is a formulation scientist with an advanced degree (Ph.D./Pharm.D., or M.S. plus experience) in pharmaceutics, pharmaceutical chemistry, or a related field, and roughly 2–5 years’ experience in nasal/mucosal drug delivery, or an M.D./pharmacologist of equivalent experience with benzodiazepine rescue therapy. I did not retrieve the exact adopted definition; treat this as a reconstruction. The POSITA would have been familiar with (a) the solubility problem for diazepam (anhydrous solubility in water ≈ 62.9 µg/mL per the Jazz reference’s own data), (b) the ≤250 µL, ideally 25–100 µL metered-nasal-spray volume ceiling, and (c) the alkyl-glycoside absorption-enhancer class (the Aegis/“Intravail” technology).

4. The threshold issue that controls the whole analysis: which references are available?

The ’664 claims differ from the already-adjudicated ’876 claims mainly in narrowing to a species/specific numbers: a defined mass (1–20 mg), a narrower tocopherol band (45–85%), a narrower alcohol band (25–40%), and the species dodecyl β-D-maltoside (rather than “an alkyl glycoside”). That narrowing is precisely what makes the priority question decisive.

Why this matters: the PTAB found in IPR2019-00451 that the ’876 claims were not entitled to the ’558 provisional date for want of written description of “alkyl glycoside,” which made Gwozdz and Cartt ’784 § 102(e) art and led to invalidation of all ’876 claims (affirmed without opinion, Fed. Cir. Rule 36, Oct. 7, 2021). Neurelis now argues the opposite for this family: because the ’414/’786/’664 claims recite the specific compound that is named in the SIGMA-catalog pages incorporated into the ’558 provisional, priority is perfected. Examiner Milligan accepted essentially that reasoning for the ’414 and ’786 (the ’414 Reasons for Allowance state that “Gwozdz [PCT] (WO 2009-120933) does not qualify as prior art”). But the Acting Director’s January 9, 2026 informative decision (declining discretionary denial in IPR2025-00464/-00465/-00466) found this directly contradicts the PTAB’s earlier determination and raises “material error.” Neurelis also seeks to disqualify Gwozdz independently via the § 103(c)(2) JRA certificate of correction (sought 12/15/2025; certificate attributed to 2/10/2026).

So there are two coherent § 103 cases. I give both.

5. Ground 1 (primary): Gwozdz PCT in view of Meezan ’962, further in view of Cartt ’784

(This is the exact ground that prevailed on the ’876 patent and that Padagis is running on the ’546/’414/’786; it has never been run against the ’664 because the ’664 is not in an IPR. Availability of Gwozdz/Cartt ’784 presupposes the ’664 loses the 2008 date and the JRA correction is ineffective.)

Element chart

Motivation to combine (why a POSITA would have done it):

1. Common field and common problem. All three address the same problem identified by the ’664 specification itself: delivering enough poorly water-soluble benzodiazepine, in a small enough nasal volume, fast enough to treat a seizure. The PTAB credited that Gwozdz, Meezan ’962 and Cartt ’784 are “naturally combinable” because they “all recognized that there are benefits to administering active agents intranasally” and “sought to provide solutions to these difficult problems” (IPR2019-00451 FWD, discussing EX1041 ¶¶ 257–258). That finding is directly transferable.
2. Express lead in the primary reference. Gwozdz itself contemplates adding excipients such as penetration enhancers — a self-executing “try this known class” suggestion under KSR.
3. The secondary reference supplies the known solution to the identified deficiency. Meezan ’962 teaches that alkyl glycosides, and n-dodecyl β-D-maltoside specifically, are effective, non-irritating intranasal absorption enhancers. A POSITA seeking to raise the bioavailability of Gwozdz’s diazepam/ternary-solvent solution would look to exactly that class.
4. Cartt ’784 supplies the dosing/volume rationale and confirms liquidity. The Board credited Dr. Wermeling that a POSA “would have looked to Cartt’784 for its disclosure of dosing regimens (including dosing amounts and volume amounts) for benzodiazepines (including diazepam)” and combined it with Gwozdz’s “intranasal benzodiazepine (diazepam) in ternary solvent systems” and Meezan ’962’s enhancers (IPR2019-00451 FWD, EX1150 ¶ 170).
5. Mechanistic compatibility. Gwozdz teaches that tocopherol + alcohol is “much less irritating to the skin and/or mucous membranes than pure alcohol solutions and generally provides higher loading of a therapeutic agent” — a teaching that supplies the rationale for the specific alcohol band (25–40%) rather than a pure-alcohol vehicle.

Reasonable expectation of success: high. The PTAB already found the Gwozdz + Meezan ’962 combination yields the claimed solution with a reasonable expectation of success and that the claimed limitations “lack criticality” with “no evidence of unexpected results” (IPR2019-00451 FWD). Adding dodecyl maltoside to an already-solubilized diazepam/tocopherol/alcohol system is an optimization of an excipient, not a new mechanism.

Range/limitation chart for the numerical limitations (Peterson/Woodruff/Aller): 45–85% sits squarely within Gwozdz’s 30–99%; 25–40% overlaps Gwozdz’s 1–70% (and its ethanol 10–30%/1–40% sub-ranges); 0.01–1% (w/v) of a maltoside enhancer overlaps Meezan’s taught enhancer loadings. Where the prior art discloses a range that encompasses the claimed range, the burden shifts to the patentee to show criticality — a burden the PTAB found Neurelis did not carry on the ’876 patent.

6. Ground 1A (dependent claims)

To the extent claims 15–18 recite only a result (purity, uniformity) rather than a structural difference, the Federal Circuit line that a newly discovered property of a known composition does not confer patentability, plus the “no criticality/no unexpected results” finding in IPR2019-00451, gives the challenger the better of the argument.

7. Ground 2 (fallback): if the ’664 claims are entitled to the 3/28/2008 date

Then Gwozdz PCT and Cartt ’784 fall out entirely as prior art (both post-date 3/28/2008), and the § 103(c)(2) JRA correction becomes belt-and-suspenders. The case must rest on art that is § 102(b) regardless:

Sonne (US 6,193,985) + Meezan ’962 + Jazz US 2008/0070904 (+ optionally the 1988 SIGMA catalog).

• Sonne discloses tocopherol-containing compositions for delivery (including nasal) of sparingly water-soluble agents such as diazepam, and co-solvents including ethanol and benzyl alcohol for optimizing bioadhesion, sprayability, and viscosity (IPR2019-00451 record, EX1013).
• Meezan ’962 supplies the alkyl glycoside, including n-dodecyl β-D-maltoside.
• Jazz ’904 supplies the intranasal-benzodiazepine (diazepam/clonazepam/lorazepam), fast-onset, low-volume spray frame of reference.
• Motivation: optimizing an intranasal route for a drug already known to need a non-aqueous vehicle, using a known enhancer to raise bioavailability, with known preservative/co-solvent alcohols.
• Weakness: a Sonne + Meezan ground was presented and rejected in an earlier IPR petition (per the Gizurarson declaration, EX2012, in IPR2019-00451: “Ground 1: Sonne [EX1013] and Meezan … was rejected by the Patent Trials Appeal Board in a separate inter partes review petition filed by Aquestive”), and Sonne’s alcohol disclosure is thin (some formulations at 5% or 7.5% (w/w) ethanol; Sonne says co-solvent addition “is less desired, since solutions of this kind tend to be irritating”). This ground is materially weaker than Ground 1.

8. What would defeat the obviousness case (counterarguments the patentee will press)

1. Priority. If the ’664 claims keep the 3/28/2008 date (via the SIGMA-catalog incorporation-by-reference theory accepted by Examiner Milligan for the ’414/’786), Gwozdz and Cartt ’784 are not prior art at all, and Ground 1 collapses.
2. § 103(c)(2) / JRA. Even if Gwozdz is § 102(e) art, Neurelis will argue the JRA safe harbor disqualifies it (common/overlapping inventors: Gwozdz, Loxley, Mitchnick appear on both; Maggio appears on Meezan ’962). Note that § 103(c) only removes art qualifying “only under” § 102(e)/(f)/(g) — so it can knock out Gwozdz PCT (published 10/1/2009, after the critical date) but not Meezan ’962 (published 3/2/2006, § 102(b)). The scope and legal effectiveness of the certificate of correction (whether a JRA statement can be added post-issuance, and whether it was in effect “on or before the date the claimed invention was made”) will be contested. I could not retrieve the certificate’s four corners.
3. In re Riggs / provisional support. Neurelis’s central briefing point is that the only motivation evidence the petitions use — Gwozdz’s “penetration enhancer” language — was not disclosed in Gwozdz’s own provisional (61/040,281), so that subject matter is not prior art at all: In re Riggs, 131 F.4th 1377, 1384 (Fed. Cir. 2025) (“the portion of the application relied on … as prior art must be supported by the provisional”). If that argument lands, Ground 1 loses its motivation to combine.
4. Genus-to-species selection. Neurelis argues the art discloses “legions of alkyl glycosides” and that selecting the C12 species “absent hindsight … would have required an incredible stroke of fortune,” invoking In re Kuban, 561 F.3d 1351, 1359 (Fed. Cir. 2009) as quoted in its papers — reported literally here; the citation is evidently to In re Kubin, 561 F.3d 1351 (Fed. Cir. 2009). Padagis counters that Meezan expressly names n-dodecyl β-D-maltoside, which neutralizes the “darts at a board” point.
5. Criticality / unexpected results. Neurelis will argue the specific 45–85% / 25–40% / 0.01–1% combination is critical because it delivers ~96% of IV diazepam bioavailability (the ’852 patent’s method claims recite exactly that). The counter is that (a) that property is claimed in a different patent with method claims and is not commensurate with the ’664 composition claim’s scope; and (b) the Board already rejected this line for the ’876.
6. Secondary considerations. The ’876 FWD found the objective-indicia evidence “weak, at best,” with no nexus and no unexpected results, and rejected long-felt need; the Federal Circuit affirmed (Rule 36). Valtoco’s commercial success and orphan-drug exclusivity exist, but nexus to the claimed composition (as opposed to the device, the intranasal route, or the rescue-care model) is contested.
7. “Consisting of” vs. “comprising.” If the corrected claim 1 uses “consisting of” (as the Strides complaint at ¶ 22 paraphrases: a “pharmaceutical solution for nasal administration consisting of” those components), the claim is narrower and the challenger must show the art teaches the exact closed set — which is precisely what was done for the ’876 claim’s “consisting of” language. If it remains “comprising” (as the 1:25-cv-00821 complaint charts it), the claim is broader and easier to invalidate. This unresolved discrepancy materially affects the difficulty of the § 103 case and should be resolved against the corrected printed claim.

9. Explicit uncertainties (carried forward and updated)

1. Corrected claim text of the ’664 is still unverified. The only new information is purpose (a Joint Research Agreement statement to invoke pre-AIA § 103(c)(2)), not the corrected wording.
2. The ’664 is not in any IPR, so no petition has yet charted these claims. Every element chart above is my construction from the ’876 FWD, the IPR2025-00464/-00465 petition summaries, and the Delaware complaint.
3. Prior-art “section” was reconstructed. The Google Patents text supplied omitted the (56) citations and the claims; my reference list is drawn from the retrieved front-page (56) capture and the IPR record.
4. Document-identity inconsistencies I am reporting literally, not correcting: the Padagis petition/exhibit headers recurrently render IPR2025-00464 as involving “U.S. PATENT 8,895,546” (which is the ’546 patent, not the ’664); the ’414 patent appears variously as 11,214,414 (paragraphfour, Neurelis press release) and 11,241,414 (Google Patents, the 1:25-cv-00821 docket); and Neurelis’s brief cites “In re Kuban, 561 F.3d 1351” (evidently In re Kubin).
5. Source-date caveats. Several retrieved sources self-report update dates after the April 26, 2026 working date, and the Strides complaint content quoted in the earlier litigation summary (filed 7/31/2026) is later than the working date. I am not reconciling those timestamps.

10. Bottom line

• Strongest § 103 case: Gwozdz PCT (WO 2009/120933) + Meezan ’962 (US 2006/0046962) + Cartt ’784 (US 2008/0279784). Every limitation of ’664 claim 1 is disclosed or suggested; the numerical ranges overlap or are encompassed; motivation is supplied by a common field, Gwozdz’s own enhancer suggestion, Meezan’s express teaching of n-dodecyl β-D-maltoside as an intranasal enhancer, and Cartt ’784’s dosing/volume teaching. This identical combination already invalidated the parent ’876 patent’s claims (PTAB FWD 8/6/2020; Fed. Cir. Rule 36, 10/7/2021), and the PTAB found the claimed ranges lacked criticality with no unexpected results. Confidence: high on the technical merits; contingent on the two threshold issues below.
• The case rises or falls on two threshold questions, not on the chemistry: (i) whether the ’664 claims are entitled to the 3/28/2008 priority date (if yes, Gwozdz and Cartt ’784 are not prior art at all, and the fallback is the materially weaker Sonne ’985 + Meezan ’962 + Jazz ’904 combination that has previously been rejected); and (ii) whether the § 103(c)(2) JRA certificate of correction is effective to disqualify Gwozdz PCT. Both are live, both are contested, and both are flagged in the previously generated sections.
• Overall confidence that at least claim 1 is invalid as obvious: moderate-to-high, conditional on Gwozdz PCT qualifying as prior art. If Gwozdz is knocked out on either ground, the obviousness case drops sharply and rests on the weaker Sonne-based combination.